Minority Report-style 'Predictive Policing' Arrives at the LAPD
Has the future of police work arrived? You might think so if you had read an article that appeared in the August 21 edition of the Los Angeles Times. Running under the headline “Stopping crime before it starts,” the article described recent advances in what is known as “predictive policing,” in which complex mathematical formulas are used to forecast when and where crimes are likely to occur.
Similar technology is already used to predict consumer behavior. The article, by Times writer Joel Rubin, tells of the decision made several years ago at Florida Wal-Mart stores to lay in an extra supply of strawberry Pop-Tarts in advance of an approaching hurricane. The numbers had been crunched, Rubin writes, and they produced an undeniable fact: “When Mother Nature gets angry, people want to eat a lot more strawberry Pop-Tarts.”
Some scientists believe this same sort of number crunching can be used to predict the behavior of criminals, regarding them as they might a flock of B.F. Skinner’s pigeons. UCLA anthropologist Jeff Brantingham is enthusiastic about the future of predictive policing, and he and his academic colleagues are working with the Los Angeles Police Department on an experiment designed to demonstrate its potential. Brantingham dismisses those who might be skeptical of his project’s value. “The naysayers want you to believe that humans are too complex and too random -- that this sort of math can't be done,” he says. “But humans are not nearly as random as we think. In a sense, crime is just a physical process, and if you can explain how offenders move and how they mix with their victims, you can understand an incredible amount.”
As one must always assume whenever government agencies team up with the lab-coat-and-sliderule crowd, there is money at stake here, specifically yours and mine. Rubin reports that the LAPD is considered a front-runner among big-city police departments competing for a $3-million grant from the U.S. Justice Department that will fund a multiyear evaluation of predictive policing. LAPD Lieutenant Sean Malinowski, who heads the department's crime-analysis unit, was recently in Washington to brief Attorney General Eric Holder on the involved concepts and technology (and presumably to get a leg up on the competition for that $3 million).
The photograph that accompanied Rubin’s story shows Lieutenant Malinowski sitting before a futuristic-looking array of computer and television monitors, from which he apparently hopes to divine where L.A.’s next outbreak of crime will occur. He envisions police officers equipped with computers to which will be streamed the latest crime information along with the addresses of paroled ex-convicts living in the area.
All of which sounds fine until you realize that a cop who spends too much of his work day staring into a computer screen will spend too little of it looking out the window of his police car for the criminals he has been so sophisticatedly equipped to detect. When I was a rookie cop in the early ‘80s, each day in roll call I was handed a copy of what was called the Daily Occurrence report, known to street cops of the time as the D.O. sheet. Most of the veteran cops I worked with gave it a glance and took note of particulars like vehicle descriptions and license numbers of cars used in crimes, but they had little use for it otherwise. They already knew where the area’s crimes were most likely to occur from having handled the calls themselves and from talking with their colleagues. They also knew where the parolees lived, which, then as now, was most often not where their parole records indicated. More important, they knew when to jump out of the car and put the grab on somebody they could see was up to no good.